Civil Engineering Reference
In-Depth Information
Whenever possible, in two-way ANOVAs, interactions of simple factors and
their possible in
uence on the dependent variable (SM in this case) will be studied.
The independent variables in a two-way ANOVA are called factors. The idea is
that there are some variables, factors, which affect the dependent variable. Each
factor will have two or more levels within it, and the degrees of freedom for each
factor is one less than the number of levels.
ANOVA procedure will be described very brie
fl
fl
y; herein further details of the
fundamentals of ANOVA can be found in [
13
].
Below, the results obtained from several ANOVAs are summarized. The anal-
yses were done for some experimental programs from the shear database analyzed
in the previous sections. In this way, the ANOVA was done each time for series
with the same geometry and cross-section and other constant variables, to have an
ANOVA as robust as possible.
8.7.2 Brite Series 1
—
In this section, the beams from Series 1 of the Subtask 4.2
Trial Beams in Shear of
Brite-Euram project [
5
] are going to be analyzed. This Serie 1 was carried out in the
Technical University of Braunschweig (UBS) by Rosenbusch J. and Teutsch M.In
Table
8.15
the constant variables within the same Series are indicated. In
Table
8.16
, the input data for the ANOVA analysis are presented, differencing the
factors (independent variables) of the dependent variable (safety margin, SM) for all
these subsections. In this case the three factors are: f
R3
, A
sw
/s and Code, having the
factor f
R3
four levels and, the other two factors, three levels. So, the number of
complete cases is:
4
1
=36complete cases
The ANOVA Table
8.17
decomposes the variability of safety margin on shear
(SM) in contributions due to several factors. Since the sum of squares Type III has
been chosen, the contribution of each factor is measured by removing the effects of
other factors. P-values, test the statistical signi
of levels)
№
of factors
=3
2
complete cases =(
×
№
№
cance of each factor. Since six
p-values are less than 0.05, these factors have a statistically signi
cant shear on SM
on a 95.0 % con
dence level.
Table 8.15
Constant
variables between the beams
within the same series (Brite
Series 1)
Constant variables
Cross-section 200
×
300 mm: Rectangular reinforced (no
prestressed) beams
ρ
l
= 3.56 %
a/d = 3.5
3 point test
Search WWH ::
Custom Search